CBTC is steadily gaining in popularity as the train control system of choice for both new and existing metro lines, and it is starting to be adopted by some mainline railways for their busiest commuter lines. The appeal is the ability to automate a metro line and thereby drive up capacity, while at the same time reduce costs and increase flexibility if the decision is made to eliminate drivers.

Until now Japanese railways and metros have eschewed CBTC in favour of domestic technology and systems even though one domestic supplier, Nippon Signal, has developed its own CBTC system called Sparcs which has been in operation on Beijing metro's Line 15 since December 2011. Nippon Signal recently won contracts to install Sparcs on Line 8 of the Delhi metro and the new 23.6km Gimpo Urban Railway in Korea.

This could all change following JR East's decision to allow Thales to install CBTC on its busy Joban commuter line in Tokyo. JR East says it is focusing on radio-based train control systems to simplify wayside equipment and is keen to compare its own Atacs system, which will be introduced on the Saikyo line in 2017, with CBTC when this comes into service on the Joban line in 2020. The first Atacs system was put into operation on the Senseki line in Sendai in October 2011 and has similar functionality to ETCS Level 3 or CBTC.

JR East says Atacs has proved to be even more reliable than expected in its first two years of operation in Sendai with a system reliability rate of 99.99999%, no location recognition errors in either refreshing a train's location or switching between Atacs and ATP, and no radio quality handover errors. Finally, JR East says it achieved a failure time of just 6 seconds. It will be interesting to see if CBTC can match these impressive figures.

In London the priority is to increase capacity as Mr David Waboso, London Underground's (LU) capital programmes director, explained to delegates at Global Transport Forum's recent CBTC conference in London. "One of the biggest problems facing metros is that they are leaving passengers behind," Waboso said. "We will increase capacity line by line by 20-30% without a huge investment by installing CBTC - it is the core of our strategy to increase capacity - but what is stopping us going to 50 trains/hour, because that is what is needed?"

While nobody yet has devised a way to get to 50 trains/hour, and it is difficult to see how they would, LU is already achieving 30 trains/ hour on the Jubilee Line and 33 trains/hour on the Victoria Line with CBTC. Speaking at the Irits summit in Berlin last November, Mr George McInulty, programme director for infrastructure with Transport for London, said with 10 to 15 more trains and "very aggressive dwell-time management" it would be possible to achieve 36 trains on the Jubilee Line.

Increasing capacity is the prime objective for LU's project to upgrade its so-called Sub-Surface Lines network comprising the Circle, District, Hammersmith & City, and Metropolitan lines, which account for 40% of the LU network and carry 25% of its passengers.

The project started life in 2003 under the former Metronet PPP concessionaire, with Bombardier supplying 191 eight and seven-car trains to replace the three types of train in operation, and Invensys resignalling the 310km network. When Metronet collapsed in 2007, LU continued with the train replacement contract, but cancelled the signalling contract as the project was deemed too expensive.

LU re-scoped the project and issued a new tender. It took the right approach to what it describes as the world's largest resignalling project by taking two years to consult other metros and select a contractor. Bombardier was awarded a £345m contract in June 2011 with the objective of completing the work by 2018. Bombardier was chosen because of its track record with installing CBTC systems on metros and peoplemovers around the world. Its CityFlo 650 system is in use on two lines Madrid for example.

However, two and a half years into the project, and despite Bombardier and LU staff being located in the same building to aid transparency and openness, and to ensure good integration of all the functions, the two parties have agreed to terminate the contract. Apart from being a major setback for LU and a serious blow to Bombardier's prestige, LU is determined to relet the contract and complete the rollout by 2018, which seems a tall order.

While both parties are being very tight-lipped about the reasons for cancelling the contract, it seems likely that the complexity of the project got the better of Bombardier and perhaps CBTC itself. Most CBTC installations are on point-to-point metro lines, but the SSL network is bedevilled with numerous flat junctions and inter-operation with other LU lines and mainline trains. It remains to be seen whether another contractor can succeed where Bombardier failed.